The overall survival impact of prophylactic cranial irradiation in limited-stage small-cell lung cancer: A systematic review and meta-analysis

Background: Prophylactic cranial irradiation (PCI) for limited-stage small-cell lung cancer (LS-SCLC) patients has become more controversial. Since the publication of the systematic review by Aupérin et al. in 1999, no randomized controlled trials regarding PCI in LS-SCLC have been completed. The aim of this study was to systematically review and meta-analyze the effect of PCI on overall survival (OS) in patients with LS-SCLC. Methods: A systematic search was conducted in the databases of MEDLINE (PubMed), Embase and the Cochrane library. Only studies that reported an adjusted hazard ratio (aHR), indicating the effect of PCI versus no PCI on OS (adjusted for confounders) in patients with LS-SCLC were included for critical appraisal and meta-analysis. A pooled aHR estimate was calculated using a random-effects model. Results: Pooling of 28 retrospective studies including a total of 18,575 patients demonstrated a significant beneficial effect of PCI versus no PCI on OS with a pooled aHR of 0.62 (95% CI: 0.57–0.69). Substantial heterogeneity of reported aHRs among studies was observed (I2 = 65.9%). Subgroup analyses revealed that this heterogeneity could partly be explained by study sample size. The pooled aHR among 7 versus 21 studies with a sample size of > 300 versus ≤ 300 patients was 0.79 (95% CI: 0.64–0.97) versus 0.56 (95% CI: 0.46–0.69; p < 0.001), respectively. Conclusions: This meta-analysis demonstrates a significant beneficial effect of PCI on OS in patients with LS-SCLC. Larger studies reported a milder beneficial effect, possibly due to a decreased risk of model overfitting. Serious risk of selection and confounding bias were of concern due to the lack of prospective trials. These results support the role of PCI in standard clinical practice in patients with LS-SCLC while awaiting results of prospective trials on alternative strategies.

Abstract licence: CC BY-NC-ND

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.